In this presentation we will discuss the benefit of using an optimal estimation retrieval framework, which provides consistence among the retrieved cloud variables and pixel-based uncertainty estimates. It also enables the homogeneous generation of cloud property datasets composed of different satellites sensors such as AVHRR, MODIS and AATSR. The strengths and weaknesses of the Cloud_cci datasets will be analyzed. Furthermore, we will highlight examples of evaluations and analyses of the three-year spanning prototype dataset that has been processed for AVHRR and MODIS in the project. The data was confronted with other well-established satellite-based datasets and ground observations. It has also been used already in comparisons with regional and global climate models and reanalysis. Finally, we will give an outlook on the three decades spanning, final dataset, which is soon to be generated.